Project Summary Promoting recovery of motor function in stroke is an essential medical and economic target. Despite significant advances in activity- and technology-based interventions, the field still fails to achieve the same level of recovery in all patients. This failure could be due to the delay in administering massed practice training until patients gain significant volitional movement, which can be weeks after stroke. Maladaptive plasticity could occur during this time and challenges the therapeutic effect of rehabilitation interventions. Exoskeleton Walking Robots (ERs) can provide assistance during walking training, and allow administering high dosage training early-on after stroke The primary hypothesis of this proposal is that the earlier the commence of high-dosage gait therapy, the better is the chance of inducing plasticity and recovery. In aim 1, investigators propose to compare the effect of progressive, long duration (30 hours over 10 weeks), ER assist-as-needed walking training (ERASSIST) versus standard of care (SOC) administered during the acute phase of a stroke on functional recovery and neuromuscular activation. Participants will be randomly assigned to each group, and investigators will examine the effect of each intervention on functional recovery and modulation of neuromuscular activations during walking. They hypothesize that early administration of intensive walking training using ER assistance can lead to significant increase in dosage, and a renormalization in walking function and neuromuscular activation during walking. In aim 2, investigators will acquire longitudinal measures of structural and resting-state functional MRI connectivity at 4 time-points: within 1 month of stroke, before and after 10 weeks of ERASSIST or SOC intervention, and at 6 months post stroke. Data will be used to model prognoses of recovery and compare the effect of ERASSIST and SOC interventions on improving the prognosis of recovery. In aim 3, investigators propose to study reorganization in cortico-muscular connectivity (CMC) post-stroke and over the course of functional recovery and compare it to gender- and age-matched healthy controls. Static CMC will be evaluated using transcranial magnetic stimulation to record ipsilesional primary motor cortex excitability and recruitment curve slope. Dynamic CMC will be explored using mobile imaging (EEG) and wireless EMG techniques to measure cortico-muscular coherence during walking. The objective of comparing stroke and healthy controls CMC data is to understand if the reorganization of CMC is an adaptation to stroke or a re-normalization to a healthy pattern of connectivity. Investigators assume that participants with high functional recovery will show a re-normalization of CMC toward more similarity with age- and gender-matched healthy controls. This research will have a significant impact on ?determining optimal combination and ?dosing? of individual interventions to improve, and possibly accelerate recovery following injury? by ?examining the underpinnings of plasticity, adaptation, and response to injury and rehabilitation intervention.?